I wanted to pass on a few comments regarding the short editorial by Mark DeSorbo (“All talk, no action” CleanRooms, July 2000). Briefly, I have two concerns: one minor and one a bit more substantial. And the two are related.
First, this is not new. I have been involved with this industry for over 20 years and for that entire time, there have been suggestions and concerns that cancer rates among fab production people are higher than normal. And throughout that interval, there have been a number of attempts to answer the questions that Mark has raised. Unfortunately, none of them arrived at any definitive conclusions.
Second, the line under the editorial title asks: “How many health experts does it take to search for cancer risks in a semiconductor fab? Moreover, how long does it take to assemble a panel of said professionals?” The implication is that people are dragging their feet, engaging in delay and somehow fighting a holding action against the search for truth. Unfortunately things are a bit more complicated.
The reasons for the apparent failure of the previous study groups are numerous.
The first is that cancer (unchecked and runaway cellular growth) is a disease that is caused by a number of apparently different factors. These causes range from genetics (often recessive traits) to environmental factors: what we wear, apply to our skins, breathe, eat, drink or otherwise swallow, or expose ourselves to in any number of creative ways.
If there are a number of possible causes and these factors can interact with each other, the difficulty in scientifically explaining a phenomenon increases exponentially. Compounding this problem is the unfortunate fact that cancer rates are strongly correlated with age.
If you live long enough, you will probably contract some form of cancer. Should we blame this on long term environmental exposure? Or on the seemingly natural process of cellular degeneration caused by normal aging? Or does long-term environmental exposure actually initiate the apparently normal process of cellular degeneration?
When scientists encounter enough interacting unknowns and they believe they are unlikely to establish the actual cause and effect mechanism, they turn to the field of statistics.
The use of statistics in scientific investigation is an attempt to provide a prediction of how something will turn out while at the same time sidestepping the issue of cause and effect.
Honest people who use statistical analysis know and readily admit that the use of statistics can lead us only to conclude certain things. They can never prove, in any absolute sense, that those things are true. Statistics can provide us only with probabilities, not certainties.
And this brings us full circle to the issue of cancer rates among fab production people. Yes, it is suspected that some of the chemicals used are carcinogenic, especially those used in photolithography. But actually arriving at this conclusion is another matter.
First, the cancer rate among fab people is not that much higher than the population at large. Further, while our sample size appears to be large (all fab production workers), nothing is being held constant.
There are a lot of different generations of products being manufactured using very different chemistries throughout the industry. Even within a single fab producing a limited range of products under relatively constant conditions and unchanging recipes, people move from one production area to another as job positions open and as production requirements dictate.
At the first statistical sign of trouble (a blip in the incidence of cancer), the industry immediately increases the exhaust rates and begins to look for alternative materials, changing the production process and thereby confounding the very studies that are trying to answer the questions that Mark has raised.
In conclusion, looking for statistically significant levels of cancer among fab production people has been extremely time consuming and challenging. And it has been the nature of the beast itself that has led to these difficulties, not to the lack of motivation on the part of the hunters.
Ken Goldstein, Ph.D.
Cleanroom Consultants Inc.